Futures margin modeling system

Abstract

A system may be configured to generate an estimate of value at risk and may include a processor to process instructions that cause the system to generate a rolling time series of value data having a plurality of dimensions, perform rotation transform of the time series, perform variance scaling and correlation scaling on transformed time series, reverse transform the results of the scaling, and estimate of a value-at-risk for the value data.

The invention claimed is: 
     
       1. A system including:
 a processor; and 
 a non-transitory memory including instructions, which when executed, cause the processor to:
 generate a time series of value data, the value data corresponding to a transaction execution value for a product of a grouping of products, the time series of value data including a first plurality of dimensions each corresponding to a different factor contributing to dynamics of the time series of value data; 
 transform the time series of value data by rotating from a first space including the plurality of the dimensions to a second space including the plurality of the dimensions with a diagonalized covariance matrix; 
 calculate a plurality of curves in the second space by applying a filtered historical simulation based on forward-looking volatility of the time series of value data; 
 statically capture an inter-curve correlation by sampling time-aligned residuals via analysis of the plurality of curves in the second space; 
 reverse transform the time-aligned residuals using a reverse rotation into the first space; and 
 estimate a value at risk for the product of the grouping of products based on the reverse transformed time-aligned residuals. 
 
 
     
     
       2. The system of  claim 1 , where in the rotation into the second space reduces the dimensionality of the inter-curve correlation. 
     
     
       3. The system of  claim 1 , wherein the filtered historical simulation includes a removal of multivariate terms from the forward-looking volatility. 
     
     
       4. The system of  claim 1 , wherein the filtered historical simulation includes a removal of autocorrelation terms from the forward-looking volatility. 
     
     
       5. The system of  claim 1 , wherein the rotation from the first space includes a rotation characterized by a decomposition matrix determined by diagonalizing the covariance matrix. 
     
     
       6. The system of  claim 5 , wherein the decomposition matrix is determined by performing a Cholesky decomposition on the covariance matrix. 
     
     
       7. The system of  claim 1 , further including a user interface device coupled with the processor and including a display device and a user input device, wherein the instructions, when executed, further cause the processor to:
 present, to the user via the display device, at least one backtesting screen comprising a result from a test; and 
 receive, via the user input device, an indication whether a model for determining a margin requirement has passed a qualitative validation test. 
 
     
     
       8. The system of  claim 1 , wherein:
 the product includes a financial product; and 
 the value data includes pricing data for the financial product. 
 
     
     
       9. A computer-implemented method including:
 generating, by a processor, a time series of value data, the value data corresponding to a transaction execution value for a product of a grouping of products, the time series of value data including a first plurality of dimensions each corresponding to a different factor contributing to dynamics of the time series of value data; 
 transforming, by the processor, the time series of value data by rotating from a first space including the plurality of the dimensions to a second space including the plurality of the dimensions with diagonalized covariance; 
 calculating, by the processor, a plurality of curves in the second space by applying a filtered historical simulation based on forward-looking volatility of the time series of value data; 
 statically capturing, by the processor, an inter-curve correlation by sampling time-aligned residuals via analysis of the plurality of curves in the second space; 
 reverse transforming, by the processor, the time-aligned residuals using a reverse rotation into the first space; and 
 estimating, by the processor, a value at risk for the product of the grouping of products based on the reverse transformed time-aligned residuals. 
 
     
     
       10. The method of  claim 9 , where in the rotation into the second space reduces the dimensionality of the inter-curve correlation. 
     
     
       11. The method of  claim 9 , wherein the filtered historical simulation includes a removal of multivariate terms from the forward-looking volatility. 
     
     
       12. The method of  claim 9 , wherein the filtered historical simulation includes a removal of autocorrelation terms from the forward-looking volatility. 
     
     
       13. The method of  claim 9 , wherein the rotation from the first space includes a rotation characterized by a decomposition matrix determined by diagonalizing the covariance matrix. 
     
     
       14. The method of  claim 13 , wherein the decomposition matrix is determined by performing a Cholesky decomposition on the covariance matrix. 
     
     
       15. The method of  claim 9 , further including:
 presenting, to a user via a display device, at least one backtesting screen comprising a result from a test; and 
 receiving, via a user input device, an indication whether a model for determining a margin requirement has passed a qualitative validation test. 
 
     
     
       16. A system including:
 means for generating a time series of value data, the value data corresponding to a transaction execution value for a product of a grouping of products, the time series of value data including a first plurality of dimensions each corresponding to a different factor contributing to dynamics of the time series of value data; 
 means for transforming the time series of value data by rotating from a first space including the plurality of the dimensions to a second space including the plurality of the dimensions with diagonalized covariance; 
 means for calculating a plurality of curves in the second space by applying a filtered historical simulation based on forward-looking volatility of the time series of value data; 
 means for statically capturing an inter-curve correlation by sampling time-aligned residuals via analysis of the plurality of curves in the second space; 
 means for reverse transforming the time-aligned residuals using a reverse rotation into the first space; and 
 means for estimating a value at risk for the product of the grouping of products based on the reverse transformed time-aligned residuals. 
 
     
     
       17. The system of  claim 16 , where in the rotation into the second space reduces the dimensionality of the inter-curve correlation. 
     
     
       18. The system of  claim 16 , wherein the filtered historical simulation includes a removal of multivariate terms from the forward-looking volatility. 
     
     
       19. The system of  claim 16 , wherein the filtered historical simulation includes a removal of autocorrelation terms from the forward-looking volatility. 
     
     
       20. The system of  claim 16 , wherein the rotation from the first space includes a rotation characterized by a decomposition matrix determined by diagonalizing the covariance matrix.